METHODS 33 



for chemistry, enabling them in the presence of thousands 

 of chemical poisons always to fabricate the specific antidote. 

 But Ehrlich's theory is also objectionable because it renders 

 barren one of our most powerful means of biological investi- 

 gation. 



It is not mere chance that has led me to choose these two 

 examples of typical errors of method. Heredity and the 

 process by which specific antitoxic serums come into exist- 

 ence are to be for us the two fundamental phenomena of 

 biology. 



